Document Type : Research Paper
Authors
1 Department of Civil Eng., Faculty of Eng., Urmia, Iran.
2 Department of Civil Eng., Faculty of Eng., Urmia University, Urmia, Iran
3 Department of Civil Engineering, Ka.C., Islamic Azad University, Karaj, Iran.
4 Department of Biosystems Mechanical Engineering, Faculty of Agriculture and Natural Resources, University of Tehran, Karaj, Iran.
Abstract
Keywords
Main Subjects
The increasing complexity of concrete dam construction projects in Iran has created an urgent need for modernized, intelligent management frameworks capable of addressing technical, economic, temporal, environmental, and cultural challenges. Iran’s dam infrastructure plays a critical role in national water security, hydropower production, flood control, and long-term economic stability. Yet, despite the essential role of these structures, the management of large-scale dam construction projects continues to face persistent difficulties such as concrete cracking, seepage, geotechnical uncertainties, scheduling delays aggravated by climate shifts, escalating costs due to inflation and currency volatility, and environmental and safety risks in seismic regions. These issues are further intensified by local constraints, including sanctions that limit access to advanced equipment, insufficient digital infrastructure, and cultural resistance among engineering teams toward adopting novel technologies.
Global research has increasingly emphasized the integration of digital technologies—such as Building Information Modeling (BIM), Internet of Things (IoT) systems, artificial intelligence (AI)-assisted prediction models, virtual reality (VR), digital twins, and advanced sensor networks—as transformative tools for increasing efficiency and reducing risk in construction. However, the successful implementation of these technologies requires a localized decision-support framework capable of capturing both quantitative and qualitative dimensions of smart management, especially within contexts characterized by uncertainty and expert-driven judgments. Although multi-criteria decision-making (MCDM) methods and machine learning (ML) models have individually contributed to prioritization and prediction in engineering management, there has been no comprehensive, hybrid framework specifically tailored to the needs of Iranian concrete dam projects.
The purpose of the present study is to fill this gap by systematically identifying, evaluating, and prioritizing 100 factors influencing smart management of concrete dam construction across nine major criteria: technical, temporal, economic, safety, cultural, environmental, legal, supervisory, and technological. These factors were extracted through extensive literature review and semi-structured expert interviews. The study aims to answer the overarching research question: Which factors exert the greatest influence on enabling smart management in Iran’s concrete dam projects, and how can these factors be prioritized through a hybrid decision-making and predictive modeling approach?
The theoretical framing of this research integrates principles from digital construction theory, resilience engineering, and uncertainty management, positioning smart management as a multidimensional phenomenon influenced by both human and technological components. The target audience includes engineering managers, policymakers within the Ministry of Energy, project supervisors, consultants, and academics seeking evidence-based guidance for digital transformation. A pragmatic rationale supports the hybrid methodology: fuzzy MCDM is ideal for quantifying ambiguous expert judgments, while machine learning enhances predictive validity, reduces model subjectivity, and reinforces stability across rankings. This combined approach provides a more robust analytical lens than either method alone.
The study employed a structured, quantitative research design combining fuzzy MCDM and supervised machine learning algorithms. Two primary data collection instruments were developed. The first questionnaire assessed 100 factors influencing smart dam construction, grouped into nine criteria. This instrument used a five-point fuzzy Likert scale and was completed by 33 experts, including senior dam engineers, geotechnical specialists, project managers, and environmental analysts with an average of 10–15 years of experience. The second questionnaire involved fuzzy pairwise comparisons of the nine criteria and was completed by 42 experts.
The analytical phase followed a sequential multi-stage process. Fuzzy AHP was used to determine the relative weights of the nine criteria, incorporating expert uncertainty into the decision process. DEMATEL was applied to analyze causal relationships and identify criteria acting as influencers or receivers within the system. Fuzzy TOPSIS served as the primary method for ranking the 100 factors, with additional MCDM methods—WASPAS, MOORA, COPRAS, and VIKOR—used for robustness checks. Together, these tools provided a comprehensive, triangulated evaluation framework.
To enhance model reliability and stability, supervised machine learning models were applied. PLSR, Ridge Regression, and Bayesian Ridge Regression were trained using defuzzified score matrices as input variables and fuzzy TOPSIS rankings as output labels. Model performance was assessed with R² and MSE metrics. PLSR achieved the best predictive accuracy (R² ≈ 0.9977), confirming the internal coherence of the MCDM rankings. All steps of the study were conducted using Excel, Python, and SPSS tools. The methodology was fully replicable and followed accepted standards for expert-based research.
Purposive sampling was used to ensure representation of key stakeholders in dam construction. Experts were selected from major Iranian universities, dam consulting firms, and the Ministry of Energy. Participation rates ranged from 85% to 90%, and all participants completed informed consent forms. No financial incentives were provided. Data were collected over six months through Google Forms, email correspondence, and interviews. All procedures complied with ethical standards governing expert consultation studies.
The intended sample size of at least 30 experts per instrument was met and exceeded. Reliability analysis yielded Cronbach’s alpha = 0.89. AHP consistency ratios remained below the recommended threshold of 0.1. The ML models exhibited high precision and negligible overfitting, reinforcing confidence in the final prioritization.
Although fundamentally quantitative, this study incorporates qualitative expert reasoning through fuzzy linguistic scales, resulting in a pragmatic mixed-method orientation. Fuzzy MCDM addresses human ambiguity, while ML introduces an empirical, data-driven layer of predictive validation. Integrating these outputs yields richer insights that neither approach could achieve on its own, particularly within the uncertain and complex environment of dam construction.
Among the nine criteria, technical and technological dimensions were found to have the strongest influence on smart management outcomes. Key factors with the highest priority scores included: limitations on importing smart equipment due to sanctions, adoption of digital twin technology for real-time simulation, IoT-based monitoring of concrete stress and structural behavior, and BIM-enabled 3D modeling of dam components. Cultural resistance, lack of training, and weak legal frameworks were also found to significantly hinder progress. The fuzzy TOPSIS results showed a high degree of consistency across robustness checks, with an average concordance rate of 95%. Machine learning verification further revealed that technical–local challenges accounted for 30–50% of the variance in factor importance. The strong alignment between MCDM and ML outputs confirmed the stability and predictive reliability of the framework.
The study demonstrates that achieving smart management in Iran’s concrete dam construction requires targeted interventions in technical infrastructure, digital capability enhancement, and policy reform. Technical and technological readiness emerged as the most critical determinants, followed by cultural adaptation and legal modernization. The hybrid framework developed in this study offers both theoretical and practical contributions: it provides a replicable model for prioritizing smart construction factors and supports national efforts toward digital transformation in water infrastructure. Future research should expand expert samples, incorporate dynamic simulation models, and extend analyses to other dam types and water systems.
This research was supported by Urmia University, Iran.
For research articles with several authors, a short paragraph specifying their individual contributions must be provided. The following statements should be used “Conceptualization, Osamah Abdulateef Abdullah AlMusawi and Mirali Mohammadi; methodology, Osamah Abdulateef Abdullah AlMusawi; software, Shahin Rafiee; validation, Osamah Abdulateef Abdullah AlMusawi, Mohammad Kheradranjbar and Shahin Rafiee; formal analysis, Mohammad Kheradranjbar; investigation, Shahin Rafiee; resources, Osamah Abdulateef Abdullah AlMusawi; data curation, Mohammad Kheradranjbar; writing—original draft preparation, Osamah Abdulateef Abdullah AlMusawi; writing—review and editing, Osamah Abdulateef Abdullah AlMusawi and Mirali Mohammadi and Shahin Rafiee and Mohammad Kheradranjbar; visualization, Shahin Rafiee; supervision, Mirali Mohammadi and Mohammad Kheradranjbar; project administration, Mirali Mohammadi; funding acquisition, Mirali Mohammadi All authors have read and agreed to the published version of the manuscript.” Please turn to the CRediT taxonomy for the term explanation. Authorship must be limited to those who have contributed substantially to the work re-ported.
All authors contributed equally to the conceptualization of the article and writing of the original and subsequent drafts.
During the preparation of this work the author(s) used ChatGPT (OpenAI) in order to draft portions of Python code for the study. After using this tool and service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.
The data from this study are available upon request from the corresponding author.
We would like to thank the Honorable Vice Chancellor for Research of Urmia University for their support in carrying out this research.
We would like to thank the honorable referees for their valuable suggestions in improving the manuscript.
The authors avoided data fabrication, falsification, and plagiarism, and any form of misconduct.
The authors declare no conflict of interest.
)